The Beginnings of Portland Cement
Portland cement is one of many materials that can be made by making a finely-ground mixture of limestone and clay, and burning the mixture at over 1000°C. The properties of the product obtained differ widely depending on the composition of the mixture and the temperature of burning.
This site is concerned with the production of Portland cement clinker and the products that are made from it. The essential processes of a cement plant have remained constant throughout the centuries.
The original “invention” or “discovery” of Portland cement sets the standard for all subsequent cement industry history, in that it cocooned in a web of disinformation, misunderstanding and lies. In this work, the term Portland cement is applied only in the modern sense. “Portland cement as we know it” can be simply defined as a calcareous cement containing (intentionally!) significant amounts of alite (“tricalcium silicate”). This mineral is only produced when the bulk chemistry contains sufficient calcium oxide to form it, and when a sufficiently high burning temperature (>1300°C) is employed. Earlier calcareous cements were burned at lower temperatures, and their main strength-giving components were belite (“dicalcium silicate”) and various calcium aluminates. Many different cements in this category were produced from the late eighteenth century onward, responding to the needs of industrialising economies.
A product named Portland cement was patented by Joseph Aspdin in 1824 (although there are other contenders for first use of the term). Joseph Aspdin, like his father before him, was born in Leeds, and lived all his life in Yorkshire.
For my French audience: Yorkshire est en Angleterre!! Joseph Aspdin N'ÉTAIT PAS ÉCOSSAIS!! Seriously though, they do know this, so why do they continue to say that he was Scottish? I can only suppose that it is very important to them that he should not have been English. Here, as elsewhere, the objective truth is of secondary importance.
The consensus of opinion is that this product was not a “true” Portland cement, and the latter was first made commercially by his son, William, in 1842. It would appear that William Aspdin felt that if he had patented his new product under a new name, the patent would have been readily infringed with impunity by his competitors, and he therefore chose to pretend that his cement was covered by his father’s patent, and retained the earlier product’s name. At the same time, he took great pains to hide the details of his process from espionage, and even from his own employees. He invented an elaborate and fictitious track record for his product.
Because of the “smoke and mirrors” surrounding the early product, and the lack in those days of any legal requirement to describe a product correctly, it is often difficult to determine when firms first made “true” Portland cement. Early entrants into the industry were fooled by the prevailing disinformation into thinking that Portland cement must be fairly easy to make (and this mistake is still made today!), and they ended up with over-simplified processes making products that were at best of marginal quality, and often did not meet the true definition of the product at all. Earlier products, including Joseph Aspdin’s, are not treated here.
Using these definitions, the first plant making the product was William Aspdin’s at Rotherhithe from 1842. The product’s arrival in the London area caused a sufficient sensation – largely as a result of Aspdin’s hyperbolic advertising – that the main Roman cement manufacturer (J. B. White) immediately set about copying it and it was produced at Swanscombe from 1845 (see I C Johnson's description of this process). Aspdin moved operations to Northfleet (what came to be called Robins Works) in 1846. The first few plants established appear to have been:
However, at this time, even heavy industrial construction was still dominated by brick masonry, and earlier types of cement were already well-established in this market. In particular, Roman Cement (made by calcining septarian nodules) dominated the market, and most Portland cement manufacturers also made this in quantity. Portland cement was commonly regarded with suspicion, partly because the extravagance of Aspdin’s early claims soon became apparent, and partly because the tricky pitfalls that await the beginner manufacturer must have made the product seem unreliable.
In France, as early as 1817, Louis Vicat had developed a product similar to that later made by Joseph Aspdin, but with a much more scientific approach. He introduced the term “hydraulique” for cements that set under water. Throughout the nineteenth century, much of the theory of cements was developed in France rather than Britain, and today the use of the term “Portland” is avoided there. It is therefore ironic that it was in France that Portland cement first found its true market: harbour works were under way in 1840-1860 at many places around the French coast, and notably at Cherbourg, Dieppe and Brest, and great use was made of concrete, using Portland cement imported from England. Although its manufacture in France began in 1853 in the Pas de Calais, England remained the main source during the 1850s and 1860s.
The plant of Dupont and Demarle at Boulogne, like so many early English plants, had originally (1848) made a "Roman" cement, using septaria from the local Kimmeridge Clay. From 1853, they made what they called "natural Portland cement", using Chalk Marl, blended, dried, ground and briquetted before burning. It was the "mother-plant" of the Société des Ciments Français et des Ciments Portland de Boulogne-sur-Mer et de Devres(!!) and remained a large operation until it was destroyed in 1941, after which it was not revived.
This export market allowed the English plants to expand production and mature their technology, and the home market for the product began to take off from 1860 onward, rapidly overtaking that of the earlier products. Exports all around the globe remained a major part of the industry’s business until the 1890s, when it suddenly became apparent that foreigners and colonials could make Portland cement as well.
In Lesley's book, p 42, an American importer is quoted:
The process of catching up with foreign competition dominates the time period (1895-2010) covered by this site.
The Early History of Portland Cement
This subject presents many difficulties, first and foremost because the accepted meanings of the terminology have changed significantly over the years. Not least among misunderstood terms is “cement”. For more than a century, cement industry people have bemoaned the fact that in common parlance, “cement” is used to mean “concrete”.
Until recently in the “Simple English” version of Wikipedia, the article on cement said “Most often it is called concrete”!
However, the origin of the term is the Latin caementum, which meant “masonry”: a caementarius was a mason, and in English usage, this was the meaning until the mid-nineteenth century. Samuel Johnson’s Dictionary says: “that which unites; mortar”, and Joseph Aspdin’s patent for Portland Cement was “my method of making a cement or artificial stone”. However, by 1850 the “official” terminology had been changed.
The earliest history of cement must therefore be discussed in terms of the history of mortar and concrete. Clearly, mortars have been used for as long as masonry has been assembled from stone or brick, and parallel innovations have taken place in different parts of the world. Lime arose as the most common “cementing” material independently in several places. Lime is easy to make (although not economically!) using primitive equipment: some convenient source of calcium carbonate, such as chalk, limestone, marble or sea-shells, is heated to around 900ºC, converting it to calcium oxide. Water is then added, and a violent exothermic reaction takes place, converting the oxide to calcium hydroxide. Lime’s most primitive application was probably as a wash to mud walls: it coagulates fine clays and would harden and stabilize the surface. A combination of lime and cattle dung, also by a coagulation process, produces a flexible and waterproof layer similar to PVC, and this was typically the basis of “wattle-and-daub” construction. The slightly water-soluble hydroxide, mixed to a thick paste with water, was used alone or in combination with sand to coat the outside of a masonry structure (as “stucco”) to waterproof it, or to produce a more aesthetically pleasing smooth surface. Subsequently mortar came to be used to form the joint between bricks or masonry blocks. The mortar stiffens initially by the removal of water from the paste, by evaporation and by absorption into the masonry, allowing successive courses to be built up without the mortar squeezing out. Addition of sand actually assists this process, by increasing the mortar’s yield stress. Many texts suggest that subsequent hardening occurs by reaction of the hydroxide with atmospheric carbon dioxide to re-form calcium carbonate.
This argument is used by shills of the lime industry to suggest that lime is “carbon-neutral”!
However, this is not the case. Hardening occurs by the “cannibalistic” growth of calcium hydroxide crystals.
In a mixture of small and large crystals of a substance suspended in a solvent that is saturated with that substance, the small crystals tend to dissolve, and the larger ones tend to grow, in a process that minimises the enthalpy of the system. Thus the larger crystals “eat up” the smaller ones.
Calcium hydroxide (mineral name portlandite) forms long, needle-like crystals which, as they grow, form a complex, interlocking mesh of crystals of increasing length. This matrix provides the strength. Reaction with carbon dioxide only occurs if the mortar is sufficiently poorly-consolidated to allow ingress of air. Norman and even Roman lime-based structures exist that are still only superficially carbonated after many centuries.
It will be clear from the above description that pure lime mortar can only be used in systems where the structural strength is provided primarily by the stacking of the brick or stone: the mortar merely “fills the gaps” and contributes little strength. Furthermore, surfaces of the mortar, if exposed to water, tend to dissolve away. The need for a mortar that would resist water predates the need for strength, at least in places where good building stone was available. In ancient Greece and Italy, the problem was solved by addition of pozzolan to the mortar. A pozzolan is a material consisting of silica and aluminosilicates in such a physico-chemical state that it will react with calcium hydroxide to produce hydrated calcium silicates. Typically the material is in a glassy phase, produced by volcanic activity. Pozzolans get their name from the village of Pozzuoli, near Naples, and various sites in the vicinity of Mt. Vesuvius have supplied weathered volcanic ash for use in construction for millennia. “Santorin Earth” is the English name of a similar material used all around the Aegean and obtained from the volcanic island of Santorini. These materials react fairly rapidly with lime, producing an insoluble matrix that can resist water.
A further advantage of the addition of pozzolan became clear to the Romans: mortars including them are considerably stronger. This led to the development of Roman concrete, consisting of rubble, sand, lime and pozzolan, mixed to a very stiff consistency with the minimum amount of water, and rammed into place with considerable force. The now-well-known inverse relationship between concrete water content and strength and the importance of complete consolidation were understood from the outset. The result of this technology had a profound effect upon subsequent understanding: many of the Roman structures sufficiently durable to have survived to modern times used concrete, and Europeans who marvelled at these during the post-renaissance period were impelled to try to emulate this Roman achievement.
It has, of course, been the European practice to say that the Romans invented concrete. However, almost identical systems were independently developed at an earlier date both in India and China, and probably elsewhere.
Iconic structures such as the Colloseum and the Pont du Gard rely upon concrete for their longevity, but the structure that most impressed later ages was the Pantheon in Rome. This 2nd century structure has an unsupported concrete dome of 43 m diameter, a size only equalled by the dome of the Florence Duomo in 1436, and not exceeded by St Paul’s in London or St Peter’s in Rome. It was not until the nineteenth century that iron structures were able to exceed it in size. Of very sophisticated design, it remains near-perfect today. As far afield as Britain, durable concrete was used during the Roman period. Practically Britain's first Roman structure - the gateway at the port of Richborough of 43AD - contained concrete. Poor transportation meant that the volcanic pozzolans of the Mediterranean region could not be used, but artificial pozzolan was available in the form of ground bricks and tiles.
Although not necessarily glassy, the de-hydroxylation of clay minerals at around 1000ºC renders their silica content highly reactive.
It is often said that the art of concrete was lost in Europe after the fall of the Roman Empire. Certainly, the number of ambitious structures diminished for 500 years, but concrete continued in use throughout the medieval period at least as a filler in masonry structures. It would appear that the strict discipline of Roman practice was lost. Lime was partially burned; brick and tile, although still included, were not finely ground, and consolidation was poor. With this, the idea of concrete as an engineering material gradually withered away, and by the Renaissance it had disappeared altogether.
Partly, perhaps, because of relatively high cost, concrete remained a mere historical curiosity until quite recently. Innovation was driven forward by the need for water resistance. There were two distinct areas of interest: the increasing use in high-class domestic architecture of stucco, to give lesser building materials an ashlar-like appearance; and an increasing number of industrial undertakings. Stucco required a material that could be worked and trowelled like fat lime mortar, but which would set and become water-proof fairly rapidly. This was a particular concern in the wet climate of Britain. Industrial uses centred around water transportation: canals and docks for ever increasing size of shipping, where both strength and water-resistance were required in mortars and grouts. At the same time, in the Age of Reason, a scientific approach was being applied to development of better materials.
Pre-eminent among the early developments was the work of John Smeaton, the civil engineer. He had the problem of constructing the third Eddystone lighthouse. Situated on a tiny rock that is submerged at high tide, 14 km from the shore, the first (1698-1703) and second (1709-1755) both succumbed to the elements, and Smeaton was charged with constructing an indestructible replacement. His structure was of dovetailed granite blocks, and as a minor detail, he needed a water-proof mortar to point the joints. Characteristically, he embarked upon a painstaking research programme to find the best available material. At the time (1756-1759) two aspects of mortar design were already well established. One was that there existed various European sources of pozzolan, addition of which would improve water resistance: the use of trass from the Rhineland was common by then in canal work. Secondly, lime from certain sources had a distinct tendency to set (instead of just stiffening) upon curing, and were water-resistant. As early as 1570, Palladio, in his treatise, mentions using a hydraulic lime made from a limestone quarried near Padua. These limes were referred to as “water-limes”.
The term “hydraulic” is a characteristically French Hellenism, and entered English parlance only in the late nineteenth century, when French advances in lime technology became known in England. The term appears to have been used first by Vicat.
Apart from this characteristic, they were similar to ordinary limes: the limestone was lightly burned, then slaked to form a fine paste. Good water-limes were known purely anecdotally, and different practitioners had their own local favourites. They included Grey Chalk limes from around Dorking and a few other locations, and various sources of Blue Lias lime.
It should be borne in mind that the science of stratigraphic geology, as pioneered by William Smith, did not arrive until the 1790s, so individual quarries were not necessarily recognised as geologically related.
For the lighthouse, a mortar was needed that would become sound and waterproof in the interval between successive tides during which construction would take place. Smeaton gathered samples both of the lime, and the rock from which it was burned, from as many locations as he could. He subjected the limes to a simple physical test: they were gauged (i.e. trowelled with gradual addition of water) neat to a stiff consistency, rolled into a ball and placed in water, and rated on their ability to harden without breaking up. He performed a crude chemical analysis of the limestones, by dissolving them in acid. He noted that the performance of the lime in water was directly related to the amount of acid-insoluble material in the parent limestone, and the really good ones slaked very reluctantly. He finally opted for a mortar using Blue Lias lime from Watchet and from Aberthaw, and pozzolan brought from Civita Vecchia, near Naples. His lighthouse remained in use until 1876 when erosion of the underlying rock necessitated its replacement. In fact, its replacement was necessitated when the still completely rigid lighthouse began to rock on its foundations in high winds. The lighthouse itself was structurally sound, and the upper part was re-erected on Plymouth Hoe, where it can be seen today.
What we now call a hydraulic lime is made from argillaceous limestone, in which the content of clayey matter is in the range 10-20%. On burning this at 900-1000ºC, there is produced a material containing ß-dicalcium silicate (belite), the amount being proportional to the silica content of the limestone. Correspondingly less free calcium oxide is produced, so that the violence of the slaking reaction diminishes. The belite reacts slowly with water to produce strength-giving insoluble hydrates.
Smeaton took his observations no further, but they were published. The implication of his observation is that, where nature fails to supply a suitably impure limestone, it ought to be possible to make a water-lime by burning an intimate mixture of limestone and clay in the appropriate ratio. However, in engineering applications, the use of ordinary lime with trass remained the normal practice. In stucco applications, many formulations were proposed, none of them influenced by Smeaton. A number of investigators in France independently noted the relationship between clay content and hydraulicity, and at least one produced an artificial hydraulic lime as early as 1774. In England, stucco was revolutionised by the introduction of Parker’s “Roman” cement in 1796.
The use of this confusing name exemplifies the way in which people were still searching for the magic ingredient that made Roman concrete so durable. In fact, nothing like Parker’s cement was ever used by the Romans. Parker set up manufacturing at Northfleet, at the site that came to be known as Robins, where an ancient tidal water mill was available for lease. He had been engaged in lime burning there for more than ten years before he discovered, and promptly patented "Roman cement" in 1796. (See patent). The product was made by calcining septarian calcareous concretions, which occur in many types of clay. In addition to calcium carbonate, they contain large amounts of clayey material, and a little pyrite and apatite. Veins in the concretions consist largely of calcite. These were to be found at many coastal locations where the parent clay is eroded by sea action, leaving the relatively hard septaria as beach pebbles. The classic locations were in the London Clay in the Isle of Sheppey and around Harwich, in Tertiary clays in the Solent, and in Jurassic clays along the coasts of Dorset and North Yorkshire: the best, which was used by Parker, was that of Sheppey. The septaria were initially gathered from the beaches, and subsequently were dredged, in massive quantities, from the off-shore seas. The septaria were roughly broken up and burned at a relatively high temperature (not because of the bulk chemistry, which is easily burnable, but because of the gross inhomogeneity of the nodules) then ground with millstones. The resulting cement was gauged with or without sand addition and rapidly applied: setting took place in five to thirty minutes.
The chemistry of these cements usually had a silica/lime ratio greater than that of belite, so they contained little free lime, and contained a certain amount of non-hydraulic lower silicates. Because they contained no free lime, they could not be slaked, and unlike hydraulic limes, grinding was unavoidable. Their content of calcium aluminates gave them rapid set. For stucco, this made them ideal. At the expense of the relaxed pace of traditional bricklaying, it could be used as a mortar in (relatively) high-strength masonry structures, and it was famously used by Marc Brunel in his Thames tunnel. Its use in government dockyard work became so strategically important that, when doubts about the reserves of good stone began to emerge, a system of rationing and taxation was proposed. Such panics were a strong spur to develop “artificial” cements that would emulate its performance, and proposals started to emerge in the first three decades of the nineteenth century.
See also the Roman Cement page.
Thus we have two distinct compositions, used in somewhat distinct applications:
It was now generally known that the raw material for rapid-setting and/or hydraulic cements should be an argillaceous limestone, whether natural or artificial. In France, a material similar to “Roman” cement was prepared in 1796 by burning and grinding certain beach pebbles in the vicinity of Boulogne. These were septaria from the Kimmeridge Clay.
The first recorded suggestion that the “artificial” limestone could be made by inter-grinding chalk and clay to a slurry is that of Vicat in 1817: he and numerous other continental manufacturers used the technique to make hydraulic limes. Vicat’s artificial hydraulic lime remained in production into the twentieth century. In Britain, the main objective was to emulate rapid-setting “Roman” cement, and several patents were issued for formulation of low-lime slurried mixtures. Of these, the most important was that of Frost.
Frost was a major manufacturer (in fact, probably at one time the biggest) of Roman cement, and was the chief government contractor in that role. Francis suggests that his relationship with the government protected him from prosecution as he made Roman cement while Parker’s patent was still active. From before 1810, he was experimenting with artificial compositions, although he didn’t obtain a patent for his “British cement” before 1822. (See patent). He set up to make it at Swanscombe in 1825. The patent makes no mention of slurrying as a method of producing a rawmix. Frost seems to have obtained much of his inspiration from Vicat, whose works he had visited, perhaps between 1822 and 1825. His method was the classical wet process that was subsequently used with little change in the cement industry for nearly two centuries. Soft Thames-side chalk was mixed with Medway alluvial clay and water in a washmill such as was already common for preparation of clay slip in the ceramics industry. The resulting thin slurry was placed in a “slurry back”: a large shallow reservoir in which it was allowed to dry out, partly by draining through the porous base, and partly by decanting the water that rose to the top. Once it had reached the “leathery” consistency familiar to ceramicists, it was sliced out in convenient sized chunks and dried on “drying flats”: surfaces of brick or iron heated from below by a coal furnace. Once dried to adequately hard lumps, these were loaded into a lime kiln in alternate layers with coke fuel and burned. This cement remained in production alongside Roman cement until around 1850. In general, it remained regarded as a cheap but inferior alternative to Roman cement. Frost departed in 1832, selling the plant to Francis and White.
Joseph Aspdin's "Portland Cement"
Among the other, much less significant, players was Joseph Aspdin , who obtained a patent for a similar product in Leeds in 1824. This, of course, was the famous “Portland cement” patent (British Patent 5022, 1824: An improvement in the modes of preparing artificial stone. See patent).
Non-academic histories of cement often include the irritating assertion that Joseph Aspdin invented Portland cement “in his kitchen stove”. No primary historical source suggests this, nor was it technically feasible. However, James Parker himself said that he had “discovered” Roman cement by casually burning septaria on his parlour fire (see Francis, p 27), and Robert Lesley seems to have carelessly compounded the stories when researching the history of the industry. As president of the US Portland Cement Association, his account became holy writ, and PCA went on to supply articles on cement history to, for example, Encyclopedia Britannica.
His process differed from that of Frost in that he was using hard limestone, and prepared it by burning it in lump form on its own, slaking it, and adding clay and water to form the slurry which was subsequently dried and burned as in Frost’s method. All commentators agree that the patent implies a product that was, at least initially, similar in properties to “Roman” cement. Aspdin moved to Wakefield and operated at two sites on Kirkgate. His eldest son James acted as his book-keeper, and his second son William became responsible for day-to-day manufacture. This arrangement continued without drawing any attention outside Yorkshire until 1841, when William departed precipitately. Joseph issued a notice to the effect that he had taken on James as his partner, that William had left, and that the partnership would not be responsible for any debts incurred by him. Joseph retired in 1844, aged 65, and left James to run the plant on his own.
Such was the early history of “Portland cement”. The term was used by Aspdin because of a resemblance of the resulting set mortar to Portland stone.
The key attribute associated with the Portland name is light colour: the Roman cements were all of an orange/brown colour due to the amount of uncombined iron oxide present. Frost, in one of his later writings (Journal of the Franklin Institute of the State of Pennsylvania, XVI, December 1835, p 376) described Roman cement as “an article of an odious colour”.
This was a comparison often made in the use of stuccos in England. Portland stone was, and remains, the prestige building stone of southern England. It has a very pale grey colour, and a fine-grained oolitic texture (typically with a scattering of small shelly fossils) that is easily sawn, chiselled and ground smooth when fresh. Smeaton had compared his mortars to it, in terms of colour and hardness. Francis points out that William Lockwood of Woodbridge was selling a product called Portland cement in 1823. It was a conventional Roman cement, made from the local septaria. Halstead quotes an amusing 1850 book on limes and cements by Burnell, referring to “Portland Cement, as it is very absurdly called”. It is unlikely that such product names cut much ice with anyone who knew what Portland stone really looks like.
Picture:© Ben Dalton 2010, and licensed for reuse under this Creative Commons Licence. Plaque commemorating Aspdin's first venture in Leeds.
However, in 1841, William Aspdin moved to London. He started, as soon as he could, to make and sell a product he called Portland cement, and this clearly caused a sensation. Isaac Johnson, in his memoirs, referred sarcastically to the “flourish of trumpets” that was made about the new cement. In particular, an independent contractor was engaged to perform tests comparing Aspdin’s product with the current Roman cements, and this showed the product to be considerably stronger. Aspdin also over-egged his promotion of his product with extravagant claims. Most notably, he claimed that his father’s product had been used in emergency repairs of Marc Brunel’s Thames Tunnel, which was one of the most famous civil engineering projects of the day. In fact, Brunel’s diaries and record keeping were extremely meticulous and it is clear from these that only “Roman” cement was used. Aspdin’s claim was one of the many barefaced lies he told in the course of his independent career.
However, his product was clearly “Portland cement as we know it”. Today, the term Portland cement means a calcareous cement that contains tricalcium silicate (alite). This silicate is much more reactive than belite and results in rapid strength gain. To make alite, the burning temperature must be above 1250ºC (usually considerably above) and in practice it is necessary to sinter (partially melt) the material to form a clinker. Clinker had always been produced in small quantities accidentally in the production of hydraulic limes and “Roman” cement as a result of “over-burning”. Clinkered material was always discarded because it would not slake, and if ground, appeared to be un-reactive. Johnson’s account of his own independent discovery of the virtues of clinker exemplifies this: he ground both the clinker and mixtures of clinker with softer-burned material and made them into paste. The mixtures rapidly set and became warm in the usual way, but the clinker alone was still soft and cold after the usual half-hour setting period, so he gave up on it and left it. Returning a few days later, he found the clinker-only pat to be much harder than the mixtures, “moreover the colour was of a nice grey”. Clearly, the ground clinker had not the rapid-setting qualities demanded in "Roman" Cement, but it conferred strength. It appears that William Aspdin carefully combined under-burned material in his product in just sufficient quantity to meet the rapid-set requirement, with the clinker producing the extra strength.
Interestingly, Lesley (e.g. p 26) says that in the USA, several “natural cement” manufacturers, when gradually converting to Portland production, made “improved” natural cements by intergrinding the two products.
By the time Isaac Johnson, who was the manager of John Bazley White’s Swanscombe plant, had succeeded in emulating Aspdin’s product, the latter had already been on the market for three years. Johnson maintained in his memoirs that his product was far superior to that of Aspdin (a tenuous claim not borne out by independent tests at the time) and implied that it was he alone who had discovered the necessity of producing clinker. Aspdin, by contrast, always kept his processes secret and made no admissions about the degree of burning that he employed. Furthermore, he always maintained that his product was the same as that of his father and was protected by the same patent. Perhaps because of the not-undeserved contempt in which William Aspdin was held by the industry, Johnson was for a long time given the credit for the invention of “true” Portland cement by first stating the importance of hard burning. The articles on cement in successive editions of the Encyclopedia Britannica consistently attribute it to him. Nonetheless, the invention is assigned to the Aspdins by more recent, technically competent writers. Firstly, comparative tests showed that Johnson’s and William Aspdin’s products were very similar, with about the same strength, and were both considerably superior to Roman cement. It is clear that Aspdin’s cement was recognised as superior by Johnson and White’s because:
The question then arises as to whether, as William Aspdin claimed, his father was the inventor of “Portland cement as we know it”. Halstead and Skempton are both inclined to say that he was, although the production and use of clinker was developed some time after the original patent. The reasons advanced in favour of this are:
However, there are persuasive counter-arguments:
In summary, it may be said that although it is possible that “Portland cement as we know it” was first produced at an earlier time in Wakefield, it was certainly produced at Rotherhithe in 1842, and the balance of probabilities is that this was in fact where it was first commercially produced.
Spread of the technology
The increase in the production and acceptance of the product after 1842 was fitful. Its initial slow progress was probably due to the fact that in England, uniquely, its major competitor was “Roman” cement, so that it was formulated for rapid set, although this militated against its most obvious application historically, in concrete. The eventual rapid growth of the market for Portland cement parallels the growth in the use of concrete, and the latter saw its first development in France. Vicat's hydraulic limes made a slow but dependable concrete, and concrete technology in France was already 30 years old when English Portland cement arrived on the scene. It was easy to ship cement to the north coast of France from the water-side British plants, and in 1840-1860 there was considerable activity in the up-rating of the facilities at the northern French ports: particularly Dieppe, Le Havre, Cherbourg and Brest. Concrete was used, both in-place and pre-cast, in the harbour works. For example, more than 6000 tonnes (four kiln-years’ production) was sent to Cherbourg by J. B. White’s alone. This work provided the British plants with a fall-back market, allowing them to develop their business and practices in these otherwise lean years. It also provided tested examples of concrete practice to British civil engineers, and rapid growth of concrete construction took place in Britain from 1860 onward, with Portland cement as almost the sole base. The sewer system of the Metropolitan Board of Works was a pioneering project, yielding much-publicized data on the advantages of Portland cement. French historians tend to down-play the importance of the arrival of English cements in France, but here again, the arrival of both Parker's Roman Cement in 1800 and the Thames-side Portland cements in the 1840s led to a intense activity to emulate them. Both were first produced in France at Boulogne, where both Kimmeridge septaria and chalk are available. The jury of the Exposition Universelle (Paris, 1855), at which the Boulogne product won a gold medal, remarked that "la supériorité du ciment de Portland (anglais) est incontestable".
The development of Portland cement manufacture outside England followed naturally wherever raw materials and a market existed. As with some districts in England, claims that “real” Portland cement was made at an early date have to be treated with some scepticism, but Skempton gives the following:
*Note: the first plant outside England and France was at Szczecin, then in Prussia, now in Poland. Please contact me if you can fill in the gaps.
The relatively late start in the USA is significant. “Natural” cements, distantly related in mineralogy to “Roman” cements were made, originally accidentally, by burning argillaceous magnesian limestones occurring in New York state and Pennsylvania. These were subsequently produced in vast quantities, and used in many civil engineering projects, particularly the canals that were proliferating in the 1820s. Whereas English Portland production overtook “Roman” cement in the late 1860s, in the USA, Portland production first exceeded that of natural cement only in 1901. The first recorded imports of British cement into the USA were in 1868, and imports from all European sources rose to a peak of 0.5 million tonnes in 1895. Indigenous manufacture allegedly began in 1871 at David O. Saylor’s Coplay, Pa plant. His US patent claimed it was “in every respect equal to the Portland cement made in England”. A few others followed suit, but by 1880 there were still fewer than a dozen kilns, and total output was 7070 t. Following the precedent of natural cement, most production was in continuous shaft kilns, typically making 10 t/day. In the 1880s, the market for Portland cement suddenly expanded, increasing seventeen-fold in the period 1878-1888: throughout this period nearly 90% was supplied by imports. Indigenous production lagged behind, but subsequently increased phenomenally: peak growth was in the decade 1893-1903, which saw a 38-fold increase, from 0.101 to 3.811 million tonnes. It overtook British production in 1902, was world leader in 1904, and accounted for 50% of world output in 1911. This, of course, corresponds exactly with the development of rotary kilns. The first successful rotary kiln came into production in 1890, and the technology rapidly displaced static kilns, with most of the many new entrants to the business buying rotary systems “off-the-peg”. This period is described in great detail in Lesley's book. The US cement trade was meticulously quantified, and the following chart summarises developments in its “heroic age”.
The period of US leadership in cement and concrete technology resulted throughout the developed world in an increased awareness of the potential uses of cement, and there followed a period of 70 years in which there was an increase, initially rapid, in underlying demand. Parallel with this was a steady increase in the quality expectations of users. A geographical pattern of demand developed - because Portland cement is much easier to use year-round in warm climates, a higher per capita demand developed in the more southerly parts of Europe and North America.
Despite a growth in underlying demand, Portland cement was (and still is) affected by the economy in general in a particularly exaggerated manner. Like other industries, consumption is affected by the resources available in a varying economy. But in addition, because the cement industry is tied to many long-term building projects, it is affected by "business confidence". The result is a cycle of demand for cement that somewhat precedes the general economic cycle, and has a considerably larger amplitude. The resulting "boom-and-bust" market had a profound effect upon the history of the British cement industry. The effect on manufacturing technology is discussed in the "Trends" section.
End-usage of cement became progressively more industrialised. This was characterised by:
The rise of a professional user base forced an improvement in the quality of cement, and also the development of a multiplicity of different cement types - not all Portland-based. Among these, in rough chronological order were:
The complexity of the product mix, and the overall market in Britain expanded steadily in the post-WWII period, with brief interruptions in recessions, up to a peak in 1973, when the UK made 21.6 million tonnes of cement, and the Irish Republic a further 1.6 million tonnes.
Developments after 1970
The fall in demand for cement in Britain after the 1973 peak has been as consistent as the rise before that date. The initial downturn resulted from the major recession and loss of confidence that followed the escalation in world energy prices. The fall was consolidated by the rise in cement manufacturing costs and prices due to higher energy costs, which made the product less competitive with other systems.
The market in Ireland historically roughly followed the British pattern, but the post-1973 period has been complicated by the effects of the "celtic tiger" bubble. The result has been the establishment of a more efficient industry than that of Britain, and a somewhat higher per capita cement consumption, given a normal economy.
More recently, concern about CO2 emissions has added to the need to reduce the amount of expensive Portland clinker in concrete mixes. The post-1970 period is characterised by the increasing use of "cement substitutes" in concrete, these consisting mainly of ground granulated blastfurnace slag (ggbs) and pulverised fuel ash (pfa).
Special plant is required at a steelworks in order to granulate the slag. Granulation causes the molten slag to solidify in a glassy state, in which form the slag easily reacts with alkali or lime to form strength-giving silicate hydrates essentially the same as those produced by Portland cement. Without the rapid chilling of granulation, the slag crystallises into unreactive minerals. The market for granulated slag grew after 1970, and granulators started to be installed. But a massive decline in British steel production followed. As mentioned above, Portland Blastfurnace Slag Cement had a small market, mainly in Scotland, since WWI. In many other countries, it formed a substantial proportion of the market. However, by the 1970s, the user end of the cement market was sophisticated, and users preferred to add ground slag at the mixer rather than to have it delivered in a "value added" form as part of the cement. There has been a certain increase in the amount of PBFC made, but it remains only a few percent of the market.
Pulverised fuel ash is produced by power stations that burn pulverised coal. Because the mineral matter in finely divided coal particles suspended in air is released directly into an intense flame, this causes melting to occur extremely rapidly - so quickly that water and carbon dioxide in its chemistry have not yet been evolved. This causes the particle to "blow up" into a thin-shelled bubble, which subsequently freezes in this form. The larger of these "cenospheres" often contain many smaller spheres inside them. To a certain extent, pfa can act as a pozzolan and contribute to strength of a concrete to which it is added. Perhaps more important is the flowability conferred by the spherical nature of the particles, allowing lower water/cement ratio and therefore lower cement content for a given strength target. Historically, the purity of ash was a problem - tarry carbon particles contaminate it and these cause delayed setting and poor concrete appearance. More recently, techniques have evolved for better power station combustion control, producing cleaner ash, and for separating out the carbon from dirty ash. As with granulated slag, the tendency has been for users to add it at the concrete mixer. A few cement plants have produced Portland Pozzolanic Cements using pfa but these have remained relatively low-volume products.
The application of carbon-reduction undertakings in the last twenty years has resulted in a much more focussed approach to reducing the amount of Portland clinker made. Even the most efficient cement kiln, burning gas, releases 0.7 tonnes of CO2 per tonne of Portland clinker made, and future technical refinements can't make much of a dent in this. Current research is therefore focussed on developing new cement systems that can replace Portland clinker. Since Portland clinker is the subject of this website, these new products are outside its scope. Despite its currently rising production in the developing world, it is expected that Portland cement will cease to be produced in quantity by 2050.
© Dylan Moore 2010: last edit 8/10/14.